An overdamped multimodal adaptive nonlinear static procedure for seismic assessment of rc infilled frames

Abstract

The seismic assessment of existing structures is considered the fundamental step to (i) estimate the seismic capacity of the initial structure (ii) predict the collapse mechanism and the structural weakness, (iii) select the most appropriate seismic retrofitting technique and determine the improved capacity of the upgraded building. Nonlinear dynamic analysis is widely recognised as the most accurate tool to predict the seismic behaviour of structures. However, this type of analysis has a high computational cost, and it is not an approach that can be extensively applied for professional purposes yet. To provide a tool that predicts the seismic behaviour of structures with a good accuracy but with a lower computational burden, nonlinear static methods of analysis were developed. The Capacity Spectrum Method (CSM) proposed by Freeman and the N2 Method proposed by Fajfar were pioneering methods and were recommended by the American and the European seismic code, respectively. Although these methods of analysis are generally reliable for the assessment of plane frames, however they neglect the contribution of higher modes of vibration to the seismic response and do not consider the progressive reduction of the structural stiffness due to the nonlinear behaviour of the structure. To improve the level of accuracy, advanced nonlinear static methods of analysis were developed, such as the Multimodal Pushover Analysis by Chopra et al., the Displacement Adaptive Puhover by Pinho et al. and the Advanced N1 method by Ghersi et al. Despite the innovative character of these methods, however they still present shortcomings. Another important aspect regarding existing structures is the presence of infill panels. Although infill panels provide the structure with a much larger stiffness and their location and mechanical properties influence the dissipative mechanism of the structure, however they are considered nonstructural elements, and their contribution to the seismic response is neglected. This thesis aims at the development of a nonlinear static method of analysis that can accurately estimate the seismic response of RC frames, with and without infill panels, keeping acceptable computation costs. To this end, the thesis proposes a multimodal adaptive procedure named overDamped Displacement Adaptive Procedure (D-DAP). This method has been developed from the combination of the approaches proposed by Pinho et al. and by Ghersi et al. The multimodal adaptive procedure to update the load vector is taken from the first, while the method for the association of the peak ground acceleration to the displacement demand without the SDOF approximation is drawn from the second. In addition, the D-DAP is equipped with an equivalent damping to consider the increase of the energy dissipation due the cumulated damage in the structure. To this end, the value of the equivalent damping is updated at each step according to a new damping law that has been properly calibrated in this work for RC frames with and without infill panels. The accuracy of the D-DAP in the seismic assessment of rc frames was compared to that of the DAP by Pinho, the MPA by Chopra, the N2 method (EC8) and the CSM (FEMA 440). To this end, a set of 54 RC frames was designed to be representative of existing buildings with various levels of seismic deficiencies, and their seismic responses were predicted by those aforementioned methods of analysis. These comparisons showed that the D-DAP applied with the proposed damping law demonstrated an accuracy in predicting the seismic response of RC frames, with and without infills, generally higher than the other nonlinear static methods of analysis. In particular, the D-DAP provided a significant improvement with respect to the other existing methods in the prediction of the response of RC frames with infill panels.